AWSAmazon AppStream 2.0End User Computing

`aws_appstream_fleet` cost estimation

A fleet of streaming application instances. Bills per instance-hour by type × fleet size. Two stream.standard.medium instances always-on are ~$146/month.

An aws_appstream_fleet provisions the instances that stream desktop applications to users. Cost is instance-hours: the desired instance count × the per-instance rate for the fleet's instance type. Two stream.standard.medium instances (~$0.10/hour each) running continuously is ~$146/month.

The decisive cost factor is fleet type. An Always-On fleet keeps instances running (and billing) around the clock for instant access. An On-Demand fleet runs instances only when users connect, billing a lower running rate plus a small stopped-instance fee, far cheaper for intermittent use. Elastic fleets go further, provisioning per session.

Instance type sets the per-hour rate (standard, compute, memory, graphics tiers), and the desired capacity sets how many run. Over-provisioning capacity for peak concurrency that rarely occurs, or using Always-On for a workforce that connects part-time, are the common overspends.

c3x prices the fleet from compute_capacity.desired_instances and instance_type, so the running cost is visible before deployment.

Terraform example

A minimal but realistic configuration that C3X can estimate.

resource "aws_appstream_fleet" "apps" {
  name          = "design-apps"
  instance_type = "stream.standard.medium"
  fleet_type    = "ON_DEMAND"

  compute_capacity {
    desired_instances = 2
  }

  image_name = "AppStream-WinServer2019-base"
}

Pricing dimensions

What you actually pay for when you provision aws_appstream_fleet.

Dimension	Unit	What's being charged
Fleet instances	per instance-hour	desired_instances × the instance-type rate. Always-On fleets bill 24/7; On-Demand bills only while users are connected (plus a small stopped fee). $0.10/hour for stream.standard.medium → 2 instances ≈ $146/month
User fees	per user-month	A per-authorized-user monthly fee may apply for certain Windows-based fleets. Usage-based.

Sample C3X output

Two stream.standard.medium instances, Always-On (continuous):

aws_appstream_fleet.apps
└─ Fleet instances (stream.standard.medium × 2)   1460 instance-hours   $146.00
                                                  Monthly               $146.00

Optimization tips

Common ways to reduce aws_appstream_fleet cost without changing the workload.

Use On-Demand fleets for part-time users

Large for intermittent use

On-Demand instances bill only while a user is connected (plus a small stopped-instance fee), versus Always-On's 24/7 billing. For a workforce that connects part of the day, On-Demand is dramatically cheaper.

Right-size desired capacity

Proportional to over-provisioning avoided

Fleet cost scales with desired_instances. Match capacity to actual concurrent sessions with modest headroom rather than peak concurrency that rarely occurs; use fleet auto-scaling to track demand.

Right-size the instance type

Tier-dependent

Standard instances cost a fraction of graphics tiers. Reserve compute/graphics instance types for genuinely demanding applications; most line-of-business apps run on standard.

Use Elastic fleets for spiky, per-session workloads

Workload-dependent

Elastic fleets provision per session rather than maintaining a running fleet, suiting unpredictable or low-volume access patterns.

FAQ

How is an AppStream fleet billed?

Per instance-hour: desired_instances × the instance-type rate. Always-On fleets bill continuously; On-Demand fleets bill only while users are connected plus a small stopped-instance fee. A per-user monthly fee can apply on certain Windows fleets.

Always-On or On-Demand, which is cheaper?

On-Demand for part-time or intermittent use, since you pay running rates only while users are connected. Always-On for workforces needing instant 24/7 access, where continuous billing is justified by zero connect latency.

How does c3x estimate the cost?

From compute_capacity.desired_instances and instance_type, pricing the instance-hours. The running cost is shown; On-Demand savings depend on actual connection hours, modelled as usage.

Related resources

AWS

aws_workspaces_workspace

Persistent cloud desktops vs AppStream's application streaming

Estimate this resource in your own Terraform

Free, open source, no API key. C3X parses your Terraform and shows line-item cost for every resource, including aws_appstream_fleet.

Get Started Star on GitHub

← Browse all AWS resources